News & analysis
Google Is Hiring to Build AI Content for Discover
Google job listings describe engineers and a design lead building AI-generated content and a "hyper-personalized concierge experience" for the Discover feed.
By Roomer Editorial · · 4 min read
What is Google hiring for?
Google has listed a Staff ML Engineer role for the Discover Feed Recommendation team. The listing says the team "leads by recommending AI-generated content (AIGC) to fulfill user interests."
The job covers quality control, topic discovery, and content understanding for AI-generated material in Discover. The posting also covers designing automated agent workflows to produce material for what it labels proactive surfaces, pointing specifically to Discover and Notifications. The role also covers checks for factuality, freshness, coherence, and tone, plus feedback loops meant to surface engaging stories globally.
A second listing, for a Senior Staff ML Software Engineer focused on AI-generated content quality in Discover, describes similar work at a more senior level. That listing says the team uses AI-generated content across major surfaces, not just Discover.
A separate data science listing for the Discover Experience team describes Discover as one of the world's largest feed products, with more than 500 million daily active users. That team owns the feed's user-facing parts, including AI integration.
What does the design role say about Discover's future?
Google is also hiring a Principal UX lead for Discover at Director+ level, based in Mountain View or New York, with a listed U.S. base salary of $275,000 to $382,000 plus bonus and equity.
The listing describes the goal as turning Discover into a "proactive, hyper-personalized concierge experience" meant to help people make progress on their goals. It also lists a responsibility for helping users move from inspiration to action, and asks for experience building value exchanges with creators and publishers.
The role would coordinate with teams at YouTube, Shopping, Chrome, and Android, and would scale AI-enabled formats across the feed. The listing wants the hire to weigh AI capabilities without losing focus on human voices and sources.
What's already visible to publishers right now?
Search Console already has a generative AI report for Discover. It counts impressions of links to a given site inside Discover's generative AI features, broken out by page, country, and date.
The report's Pages view lists canonical URLs that supplied information for those features, which may not match the page a reader eventually lands on. Impressions are the only metric the help page describes. Search Labs experiments aren't included, and sites with few impressions may not see the report populate at all.
Whether a site's content shows up in Discover's generative AI features is controlled through the Search generative AI control, the same setting that governs AI Overviews and AI Mode. Excluding a site there keeps its links out of those generative features while leaving other sites unaffected.
This isn't new infrastructure arriving with these job listings. Back in October 2025, Google started surfacing AI-generated previews of trending Discover topics across the U.S., South Korea, and India, with each preview linking back to the original web content. Google said in August that readers would be able to steer their Discover feed by describing, in plain language, what they wanted more or less of. Robby Stein, who serves as Google Search's VP of Product, said in September the company had begun piloting a feature called Dive deeper, giving users a quick rundown of a topic along with links to further reading, kicking off with video clips.
What's still unclear?
None of the Discover job listings includes a timeline. None describes how much of the feed AI-generated content would eventually fill.
Google Careers doesn't list when these roles went up, and listings can change or close without notice. Stein said Google planned to test multiple "Dive deeper" designs in the weeks following his September post, but no broader rollout schedule has been confirmed.
Roomer Editorial's take
The honest read here, in Roomer Editorial's view, is that Google is staffing up to generate more of the feed itself, not just rank what publishers already made. That changes the math for anyone whose traffic depends on getting pulled into Discover.
It doesn't change what you should do. Discover has always rewarded distinctive, specific content over generic summaries, and that bar goes up, not down, when Google is also generating generic summaries at scale.
The practical move is to keep an eye on your own Discover generative AI report in Search Console rather than wait for Google to announce anything official. If impressions tied to AI features on your pages start climbing or dropping, that's a signal worth acting on before any company blog post confirms the shift.
The deeper point for creators building outside the big platforms: owning a site under your own name, with your own video and your own commission structure, means a Discover algorithm change is a traffic story, not an existential one. If your business model is "whatever Google decides to show," you've already outsourced your decision-making to a team that, per these listings, is actively trying to need you less.
Check your Discover generative AI report in Search Console this week and note your baseline impressions before any further changes roll out.
Source: Search Engine Journal
